The recent Data Council Conference in Austin hosted a dedicated audience focused on managing the technical demands of running massive data-driven services. Attendees came from behemoths like Airbnb, eBay, and Wayfair, as well as representatives from startups that support, accelerate, and anticipate the growth of these massive firms.
In the age of artificial intelligence, discussions naturally revolved around AI and LLMs, but this audience of data scientists predictably also delved into topics such as data governance, data culture, data supply chains, data collaboration, and data quality. As the age-old saying goes, “garbage in, garbage out,” and data quality continues to be a core aspect of every data science and AI conversation.
Popular and timely questions included:
- How do you define quality?
- To what extent does data hygiene factor into your product development cycle?
- How much time does your team spend on normalizing and/or cleansing data?
- How much scrutiny is placed on the quality and provenance of datasets?
These innovative leaders consistently need to accelerate their product development cycles, and the requirements for the types of partners they need were universal: flexible, collaborative, and fast. This necessitates a combination of ‘big iron’ and lean code for user responsiveness. Building quality data management systems also necessitates an ecosystem of high-quality data providers and service firms that can move quickly and deliver at scale.
It’s no exaggeration to say that our economic future depends on the success of firms that are reinventing travel, transport, retail, healthcare, information, and every other industry. But for these innovators to reach their potential, they need their engineering and product development teams focused on building services – and that means, more than ever, they need reliably accurate, world-class business-to-business data.